Quadratic Spreadsheet

Use this skill when a task involves Quadratic files, spreadsheet ranges, formulas, Python or JavaScript code cells, SQL connection cells, imports, or spreadsheet formatting.

Workflow

1. Confirm the Quadratic MCP server is available as quadratic.
2. If authentication is needed, call auth with action login and guide the

user through the device authorization flow.

1. Use files_read to list or identify the target file, then files_write to

open it.

1. Start with read_data action get_spreadsheet_context or outline to

understand sheets, tables, code cells, charts, images, and existing data.

1. Use targeted read_data calls for specific ranges or tables. Consolidate

independent reads with action: "batch" when useful.

1. For edits, use write_data and place new content outside occupied ranges.

Leave a blank row or column between new content and existing data.

1. After writes, read back the changed range or context needed to verify the

result.

Notes

• Quadratic tables do not support formulas or code cells inside the table.
• Code cells are not global; variables and imports are scoped to each cell.
• Use formatting tools for cell formatting rather than trying to format from

Python or JavaScript code.

• For table columns, prefer table references such as TableName[Column].